Conventionally, there is a proposal of a method for automatically creating moving picture corresponding to characters included in character strings (e.g., see Unexamined Japanese Patent Publication No. 2002-366964).
According to this method, a character string is analyzed first and a concept contained in the characters included in the character string is determined. Next, components of moving picture such as story, props (gadgets), background and music and so forth are selected. Then, a series of animation is created using the selected moving picture components.
However, the conventional method only creates a series of animation using the moving picture components related to the objects included in the character string and pays little attention to rendering (here “rendering” has a broad meaning containing both direction which means character animation, camera animation and rendering which means drawing polygon per second) such as the method of taking moving picture (i.e. camera work).
Thus, the conventional method simply arranges moving picture components which produce premeditated actions, which results in similar rendering irrespective of themes of moving pictures. For example, similar camera work applies to all types of moving pictures such as a moving picture related to landscape and a moving picture related to human action. For this reason, the moving pictures created become monotonous pictures unable to emphasize their themes.